Quality control via accelerated data with exponential distribution and power law stress-response relationship

Sabrina Luzia CAETANO[1]


§         ABSTRACT: Obtaining measurements from manufactured product reliability is very important for industrial quality control. Such control can usually be carried out in the production line environment via the so-called accelerated tests. These tests consist in the submission of certain number of units of the product to accelerated stress levels. The main objective of this work consists in the study and the proposition of a methodology of accelerated tests for quality control. We consider a exponential distribution and a power stress-response relationship. Resample procedures are considered for interval estimation. Simulated data illustrated the proposed methodology.

§         KEYWORDS: Asymptotic confidence interval; parametric bootstrap; censoring; quality control; simulation; accelerated tests.



[1] Departamento de Estatística, Universidade Federal de São Carlos – UFSCar, Caixa Postal 676, CEP: 13565-905, São Carlos, SP, Brasil. E-mail: dfln@power.ufscar.br